Data security is rapidly becoming the most important, and potentially limiting factor in the fields of data store, data security, and data processing. While the emergence of portable data, “cloud computing,” and other forms of distributed data processing and data sharing have the potential to provide truly revolutionary and paradigm shifting advances in human activity, the current inability to provide adequate levels of data security has prevented the realization of the full potential of these advances and capabilities.
In a cloud computing environment, an organization can control which resources a given party has access to by controlling access secrets, such as passwords and other authentication credentials that enable users to gain access to data. Nevertheless, in spite of the tremendous lengths to which organizations go to prevent unauthorized access to data, fraudsters still find ways to access the data. For example, fraudsters often find and exploit flaws in security and access protocols in order to gain unauthorized access to data. When this occurs, it is possible that fraudsters may find and exploit very sensitive data related to users of the data management system.
Due to these risks, traditional data management systems take further measures to prevent fraudsters who have gained access to sensitive data from making use of the sensitive data. For example, traditional data management systems typically encrypt their data so that if a fraudster gains access to the data, the data is not in a form that is useful to the fraudster without the proper encryption keys. However, resourceful fraudsters, given enough time, can still find ways to break the encryption and gain full access to the data.
Another common problem with traditional data management systems is that many traditional data management systems only encrypt data a single time. Once the data is encrypted, these traditional data management systems never re-encrypt the data. If a fraudster gains access to the encrypted data, the fraudster may have an indefinite period of time during which to break the encryption. Some traditional data management systems may re-encrypt data on occasion. However, this weak re-encryption is typically handled in an ad-hoc manner such that years may pass before an organization that maintains a data management system will decide to manually enact a re-encryption process. A fraudster that gains access to the encrypted data may have years to break the encryption before the data is re-encrypted. Eventually, the fraudster will succeed.
Furthermore, some traditional data management systems may encrypt an entire database with a single encryption key. If a fraudster breaks this single encryption scheme, the fraudster gains exploitable access to the entire data base. Other traditional data management systems may encrypt various records or rows of a database with different encryption keys. However, these traditional data management systems still suffer from the drawback that if a fraudster breaks the encryption on one portion of the database, the fraudster will be able to exploit that portion of the database to the detriment of the data management system and the users whose data has been compromised.
Consequently, there is a long standing technical problem in the data management arts in the form of a need to provide more effective data encryption procedures.